From meat packing to quantum computing: vertical integration
In the 1920s the Ford Motor Company controlled factories for rubber, glass, and metal, and of course for the production of automobiles. Interestingly, Henry Ford writes that he learned of this strategy after visiting the meat-packing factories of Gustavus Swift, a pioneer both in vertical integration as well as in the establishment of production lines.
Similarly, the Carnegie Steel Company was also vertically-integrated: besides the steel mills themselves, Carnegie also owned iron mines that delivered iron from which steel was made, coal mines to provide fuel for the factories, and even the railroads used to deliver the raw material and the finished product.
Merriam-Webster defines vertical integration as “the combining of manufacturing operations with the source of materials and/or channels of distribution under a single ownership or management.”. Whether to maximize profits or overcome supply-chain challenges, vertical integration is a popular business strategy.
Back to the 21st century, we are seeing vertical integration plays in quantum computing as well, with mergers between quantum hardware and quantum software companies. Shareholder benefits aside, is quantum vertical integration good for customers?
Proponents of vertical integration in quantum would say that it’s a good way to ensure that the software works well — indeed is optimized — to the particular characteristics. Whether it is the noise model, the connectivity, or the gate set, tightly-integrated software might be able to account for a limitation or take advantage of a special feature. Customers could conceivably have a ‘one-stop shopping’ experience, where they get everything from a single vendor.
There is also a strong counter-argument to be made. If a customer asks what is the best quantum computer to execute a particular algorithm, the answer might change over time. New quantum modalities, new hardware implementations, and new computer models come out in rapid succession. The “best” vendor today might not be the best vendor tomorrow, and customers might be concerned about committing to one particular vendor and jeopardizing their ability to port the software to another vendor. Additionally, would it be fair to expect a single company to be the best in hardware as well as software, or is there a chance that the best software does not come from the same company that brings the best hardware?
Indeed, a Dec ’21 survey by Zapata reveals that “Organizations Need Quantum Vendors But Are Concerned About Getting Locked-In.” Seeking input from some 300 enterprises the survey found that while 96% of respondents agreed that they could not successfully adopt quantum computing without the help of a trusted vendor, 73% of respondents were concerned about getting locked in with a full-stack quantum vendor, while 47% were very or extremely concerned. Amongst those surveyed, the top consideration for choosing a quantum vendor was forward compatibility, a hardware-agnostic, interoperable approach.
What do you think? Do you prefer a one-stop-shop for all your quantum hardware and software needs, or would you opt for ‘best of breed’ solutions that can adapt to changing technology and market conditions?
Originally published at https://www.classiq.io.